[OVEP] OpenVINO Development Updates by ankitm3k · Pull Request #28954 · microsoft/onnxruntime

ankitm3k · 2026-06-09T06:18:07Z

Description

This PR introduces minor development patches as listed below.

Fix the OVIR EpCtx model export fixing the changes introduced by Unify path validation in EPs, address security gaps #28725
Add test for OVIR EpCtx Export on sample model.
Update OV Toolkit version to 2026.2
Add test for OVEP Workload Type Test.

* Update operator support status for OpenVINO 2025.2 * Disable unsupported tests --------- Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com>

Sync with Microsoft ONNX Runtime - 28/08/2025

@skottmckay

* [CPU] Optimize GQA attention bias application for FP16 (microsoft#25871) ### Description When using attention bias input for GQA op with FP16, on the platforms that don't natively support FP16 math a cast to fp32 needs to be performed, and thus a temporary buffer needs to be created to store the fp32 values. The issue is that this temporary buffer was being allocated / deallocated inside of a loop for every token being processed. Refactored the implementation so that the allocation takes place only once. Phi model throughput increased by 15%. * Fixes for DynamicQuantizeMatMul and Attention3D tests (microsoft#25814) ### Description This change fixes correctness issues in two areas that were causing failures in onnxruntime_test_all: - DynamicQuantizeMatMul.WithConstantBInputs - AttentionTest.Attention3DDefault - AttentionTest.Attention3DWithPastAndPresentQkMatmul What was wrong and how it’s fixed 1) DynamicQuantizeMatMul.WithConstantBInputs - Root cause: The Kleidi dynamic quantization GEMM path could be selected even when the B scales contained values such as (zero, negative, or non-finite). That violates kernel assumptions and can lead to incorrect results. - Fix: In `onnxruntime/contrib_ops/cpu/quantization/dynamic_quantize_matmul.cc`, we now explicitly validate that all B scales are finite and strictly positive before enabling the Kleidi/MLAS dynamic path. If any scale is invalid, we disable that path. 2) Attention tests (Attention3DDefault, Attention3DWithPastAndPresentQkMatmul) - Root causes in `onnxruntime/core/mlas/lib/kleidiai/sgemm_kleidiai.cpp`: - Incorrect handling of GEMM corner cases for alpha/beta and K==0 (e.g., not respecting C = beta*C when alpha==0 or K==0). - Unnecessary or premature fallbacks for small shapes. - Fixes: - Add early-outs for degenerate sizes: if M==0 or N==0, return handled. - Correctly implement alpha/beta semantics: --------- Signed-off-by: Jonathan Clohessy <jonathan.clohessy@arm.com> * Fix MoE CPP tests (microsoft#25877) This change adds skip test for QMoE CPU tests when running on TensorRT or CUDA EP. In the QMoE kernel there was a memory overwrite bug in the accumulate part, updated that and this fixed the python tests back * [c++] Eliminate dynamic initialization of static Ort::Global<void>::api_ (microsoft#25741) ### Description Delay the call to `OrtGetApiBase()` until the first call to `Ort::GetApi()` so that `OrtGetApiBase()` is typically called after dynamic library loading. ### Motivation and Context When ORT_API_MANUAL_INIT is not defined (which is the default), the static `Ort::Global<void>::api_` has a dynamic initializer that calls `OrtGetApiBase()->GetApi(ORT_API_VERSION)` This dynamic initialization can cause problems when it interacts with other global/static initialization. On Windows in particular, it can also cause deadlocks when used in a dynamic library if OrtGetApiBase()->GetApi() attempts to load any other libraries. * Replace the templated `Global<void>::api_` with an inline static initialized to nullptr. * `Ort::GetApi()` now calls `detail::Global::GetApi()` which calls `detail::Global::DefaultInit()` if initialization is needed. * When `ORT_API_MANUAL_INIT` is defined, `DefaultInit()` returns nullptr, which will eventually cause the program to crash. The callers have violated the initialization contract by not calling one of the `Ort::InitApi` overloads. * When `ORT_API_MANUAL_INIT` is not defined, `DefaultInit()` uses a function-level static to compute the result of `OrtGetApiBase()->GetApi(ORT_API_VERSION)` once and return it. * `Ort::Global<void>` has been replaced with a non-templated type and moved inside a `detail` namespace. Since the `Global<void>` object was documented as being used internally, it is believed that these changes here are non-breaking, as they do not impact a public API. The public APIs, `Ort::InitApi()` and `Ort::InitApi(const OrtApi*)` remain unchanged. * Add `#pragma detect_mismatch` to surface issues with compilation units that disagree on how ORT_API_MANUAL_INIT is defined. (MSVC only.) --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * python GPU IO Bindings for NVIDIA (microsoft#25776) ### Description  1. A Small change to use the shared allocator in Python binding. 2. Remove the FP64 support from the EP. ### Motivation and Context  The Python GPU IO binding is necessary for performance. The change will enable the shared allocator for GPU allocation. The FP64 was using the FP32 inference—aligned WRT TRT RTX support. --------- Co-authored-by: Gaurav Garg <gaugarg@nvidia.com> * [CANN] Add a `enable_cann_subgraph` feature parameter (microsoft#25867) ### Description Add a `enable_cann_subgraph` feature parameter. this parameter controls whether graph splitting is performed and can help quickly identify issues in certain scenarios. * [EP ABI] Add OpAttr_GetTensorAttributeAsOrtValue and replace the existing Node_GetTensorAttributeAsOrtValue (microsoft#25886) ### Description Replace `Node_GetTensorAttributeAsOrtValue` with `OpAttr_GetTensorAttributeAsOrtValue`. Change the API signature to make it one of the `OpAttr` interfaces instead of the `OrtNode` interface. The original API was added [here](microsoft#25566). * Language bindings for model compatibility API (microsoft#25878) ### Description This change builds on top of microsoft#25841 , and adds the scaffolding necessary to call into this API from C++ / C# / Python. ### Motivation and Context microsoft#25454 talks more about the broader notion of precompiled model compatibility. This change is directed at app developers whose apps may want to determine if a particular precompiled model (e.g. on a server somewhere) is compatible with the device where the application is running. There is functionality in `OrtEpFactory` for making this determination, which was exposed as a C API in microsoft#25841, and this change makes the API more broadly available in other languages. ### Testing and Validation Introduced new unit test cases across each language, and verified that the API was being called and returned the correct result for the default CPU EP. --------- Co-authored-by: Aditya Rastogi <adityar@ntdev.microsoft.com> * [QNN-EP] Introduce Level1 Transformer into qnn.preprocess (microsoft#25883) ### Description - Introduce Level1 Transformer into qnn.preprocess to support various optimizations. ### Motivation and Context - This change brings in several useful optimizations such as `ConvBnFusion` and `ConstantFolding`, which are part of `TransformerLevel::Level1` and can benefit QNNEP. - The goal is to optimize the ONNX model before quantization by integrating these passes into the Python tooling workflow. * [QNN EP] Minor fix weight name missing when not valid QDQ node group (microsoft#25887) ### Description Minor fix weight name missing when not valid QDQ node group ### Motivation and Context Some quantized model failed QDQ node group validation, the weights then won't be folded as initializer. QNN EP failed to handle the dynamic weights here due to the transpose op input name look up. This change make sure we process the weights tensor before adding transposes. * Add custom ops library_path to EP metadata (microsoft#25830) ## Summary Adds EP metadata library path support to enable custom ops DLL registration with proper path resolution. ## Changes - Added `library_path` metadata key to EP metadata infrastructure - Pass resolved library path directly to `EpLibraryProviderBridge` constructor - Simplified implementation per reviewer feedback (removed virtual method complexity) - Added `#include <utility>` for std::move compliance ## Purpose Enables downstream applications (like onnxruntime-genai) to resolve relative custom ops library paths using EP metadata, improving DLL registration reliability. ## Files Modified - `plugin_ep/ep_factory_provider_bridge.h` - `plugin_ep/ep_library.h` - `plugin_ep/ep_library_plugin.h` - `plugin_ep/ep_library_provider_bridge.cc` - `plugin_ep/ep_library_provider_bridge.h` - `utils.cc` * [OVEP] OpenVINO EP Features and bug-fixes for ORT-1.23 (microsoft#25884) ### Description This update introduces multiple improvements, fixes, and feature enhancements to the OpenVINO Execution Provider (OVEP) and related components in ONNX Runtime: #### Configuration & Properties - Updated load_config mapping to act as a passthrough to OpenVINO properties. - Added support for providing layout information to inputs/outputs in OpenVINO. #### Inference & Tensor Handling - Improved OVInferRequest::SetTensor to correctly handle cached binding shape mismatches. - Added support for self-detecting on-the-fly bfloat16 → float16 conversion. - Fixed issues with input ONNX models when used with shared execution contexts. #### Model Handling & Operator Support - Fixed model copying behavior for QDQ stripping. - Updated operator support status for OpenVINO 2025.2. #### Platform & Integration Fixes - Applied multiple PSU Lora fixes and related updates. - Resolved filename confusion issues with wrapped OVIRs in EPCtx. - Enabled memory-mapped native binaries for OpenVINO 2025.3. #### Quality & Maintenance - Addressed linting issues. - Fixed coverage gaps in OVEP. - Added a new test script for OpenVINO with ORT ABI integration. --------- Co-authored-by: Ankit Maheshkar <ankit.maheshkar@intel.com> Co-authored-by: Ryan Metcalfe <107415876+RyanMetcalfeInt8@users.noreply.github.com> Co-authored-by: Klimenko, Mikhail <mikhail.klimenko@intel.com> Co-authored-by: sfatimar <sahar.fatima@intel.com> Co-authored-by: Garth Long <garth.long@intel.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com> Co-authored-by: Eric Crawford <eric.r.crawford@intel.com> Co-authored-by: jatinwadhwa921 <110383850+jatinwadhwa921@users.noreply.github.com> Co-authored-by: Vishnudas Thaniel S <vishnudas.thaniel.s@intel.com> Co-authored-by: Javier Martinez <javier.e.martinez@intel.com> * [java] Auto EP and compile model support (microsoft#25131) ### Description Java API for compile model and EP discovery APIs. Roughly equivalent to the C# version in microsoft#24604. cc: @skottmckay. I haven't quite got the CMake configured so the Java tests for the ep registration only run when the ONNX Runtime shared provider support is built, but everything else works. I expect that to be a quick fix, but I'm not sure in what conditions it should be built and how we should handle it so I don't know where/when to plumb it through. ### Motivation and Context API parity for Java. * Add error handling to extract_nuget_files.ps1 (microsoft#25866) ### Description 1. Check process exit code when running 7z.exe . Currently the errors were silently ignored. 2. Add snld20 flag to the 7z.exe commands, which is needed to be compatible with the latest 7z release. * [Fix] illegal memory access in GetInputIndices with optional inputs (microsoft#25881) ### Description Fix illegal memory access in GetInputIndices with optional inputs ### Motivation and Context When an input is optional, its ValueInfo may be nullptr. The current implementation directly calls InputValueInfo->GetName(), leading to illegal memory access. Update logic to skip optional inputs when valueInfo is nullptr . * Re-enable cpuinfo for ARM64EC (microsoft#25863) ### Description  Re-enable cpuinfo for ARM64EC build and fix `CPUIDINFO_ARCH_ARM` so it is actually used. Patch cpuinfo to support vcpkg ARM64EC build. See pytorch/cpuinfo#324. ### Motivation and Context  Fix for workaround in microsoft#25831. --------- Signed-off-by: Jonathan Clohessy <jonathan.clohessy@arm.com> Co-authored-by: derdeljan-msft <derdeljan@microsoft.com> Co-authored-by: Jonathan Clohessy <jonathan.clohessy@arm.com> Co-authored-by: Akshay Sonawane <111780983+apsonawane@users.noreply.github.com> Co-authored-by: Christopher Warrington <chwarr@microsoft.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: Ishwar Raut <iraut@nvidia.com> Co-authored-by: Gaurav Garg <gaugarg@nvidia.com> Co-authored-by: Xinpeng Dou <15529241576@163.com> Co-authored-by: Chi Lo <54722500+chilo-ms@users.noreply.github.com> Co-authored-by: adrastogi <aditya.rastogi@microsoft.com> Co-authored-by: Aditya Rastogi <adityar@ntdev.microsoft.com> Co-authored-by: qti-hungjuiw <hungjuiw@qti.qualcomm.com> Co-authored-by: qti-yuduo <yuduow@qti.qualcomm.com> Co-authored-by: Pradeep Sakhamoori <psakhamoori@microsoft.com> Co-authored-by: Preetha Veeramalai <preetha.veeramalai@intel.com> Co-authored-by: Ankit Maheshkar <ankit.maheshkar@intel.com> Co-authored-by: Ryan Metcalfe <107415876+RyanMetcalfeInt8@users.noreply.github.com> Co-authored-by: Klimenko, Mikhail <mikhail.klimenko@intel.com> Co-authored-by: sfatimar <sahar.fatima@intel.com> Co-authored-by: Garth Long <garth.long@intel.com> Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com> Co-authored-by: Eric Crawford <eric.r.crawford@intel.com> Co-authored-by: jatinwadhwa921 <110383850+jatinwadhwa921@users.noreply.github.com> Co-authored-by: Vishnudas Thaniel S <vishnudas.thaniel.s@intel.com> Co-authored-by: Javier Martinez <javier.e.martinez@intel.com> Co-authored-by: Adam Pocock <adam.pocock@oracle.com> Co-authored-by: Changming Sun <chasun@microsoft.com> Co-authored-by: mingyue <131847423+mingyueliuh@users.noreply.github.com> Co-authored-by: Edward Chen <18449977+edgchen1@users.noreply.github.com>

This reverts commit 2f1ad9d.

Revert "Sync with Microsoft ONNX Runtime - 01/09/2025"

Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com>

Handled tensors with bf16 data type and external in memory data Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com>

Backmerging with Msft commits

Co-authored-by: Preetha Veeramalai <preetha.veeramalai@intel.com>

Sync with Microsoft ONNX Runtime - 15/09/2025

Sync with Microsoft ONNX Runtime - 17/09/2025

Fix issue with creating backend for every inference iteration. The issue was caused by using std::map operator[] that created a pair with key and empty value. In Ubuntu std::map insert method won't override the key value in backend_map if key exists in map (created by operator[])

Sync with Microsoft ONNX Runtime - 26/09/2025

Sync with Microsoft ONNX Runtime - 01/10/2025

…817) * early version, it doesn't embed initializers into the proto, but then restores the metadata so OV can read them back Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> * improve code, refactor into smaller functions, run the logic when there are external initializers in memory (more than one) Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> * revert the wrongly merged code Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> * Updated the condition for the new logic based on the total size of ext initializers, comments, refactoring Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> * Update onnxruntime/core/providers/openvino/backend_manager.cc Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * make the condition less strict - 32MB threshold, move debug dump after the logic is executed, check for OV version Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> * unit test that uses ext initializers, early version Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> * used kOrtSessionOptionsDisableCPUEPFallback, cleanups, model is now over 2GB to show the proto limit (when the new logic for ext initializers is enabled, then the test passes) Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> * address code review comments Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> * Update onnxruntime/test/providers/openvino/openvino_ep_ext_init.cc Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * fix the Linux CI build, use PathString rather than wstring Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> * As agreed, disable the test as it requires OV 2025.4, while the current CI version is only 2025.2 Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> * add missing comment Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> --------- Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

…zers (#825) * check for the size of RAW data, skip if it's zero Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> * Update onnxruntime/core/providers/openvino/backend_manager.cc Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> --------- Signed-off-by: bfilipek <bartlomiej.filipek@intel.com> Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Sync with Microsoft ONNX Runtime - 10/10/2025

… from initializers might become invalid (#828) Signed-off-by: bfilipek <bartlomiej.filipek@intel.com>

Co-author: Beheshti, Nazanin

Backmerge pr Msft commits

Sync with Microsoft ONNX Runtime - 08062026

Update OV version to 2026.2.0

Sync with Microsoft ONNX Runtime - 09062026

ankitm3k · 2026-06-09T06:21:36Z

@yuslepukhin & @adrianlizarraga Please review & merge this PR update.

Copilot

Pull request overview

This PR updates the OpenVINO Execution Provider (OVEP) development stack by bumping the OpenVINO toolkit version to 2026.2, adjusting OVEP’s version gating/mapping to recognize 2026.2, fixing EPContext path validation behavior for OVIR-exported artifacts, and adding a new workload-type dynamic-options regression test.

Changes:

Update OpenVINO toolkit references (Linux docker image + Windows CI workflow) to 2026.2.0.
Extend OVEP version plumbing to include 2026.2 (new enum + capability mapping + unsupported-mode lists).
Fix EPContext external path validation base-path selection for .xml cache contexts; add workload-type switching tests.

Reviewed changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
tools/ci_build/github/linux/docker/inference/x86_64/python/openvino/Dockerfile	Bumps Linux OpenVINO toolkit package + install path to 2026.2.0.
onnxruntime/test/providers/openvino/openvino_ep_workload_type_test.cc	Adds new OVEP workload-type dynamic switching and repeat-run consistency tests for NPU.
onnxruntime/core/providers/openvino/ov_versions/data_ops.h	Adds OpenVINO version enum value `V_2026_2`.
onnxruntime/core/providers/openvino/ov_versions/data_ops.cc	Extends unsupported-mode version lists to include `V_2026_2`.
onnxruntime/core/providers/openvino/ov_versions/capability.cc	Maps OpenVINO 2026.2 builds to `V_2026_2` behavior.
onnxruntime/core/providers/openvino/onnx_ctx_model_helper.cc	Adjusts EPContext external path validation base for `.xml` cache contexts.
cmake/onnxruntime_providers_openvino.cmake	Raises minimum supported OpenVINO version to 2026.0+.
.github/workflows/windows_openvino.yml	Updates Windows CI OpenVINO download/extract to 2026.2.0.

Comments suppressed due to low confidence (1)

cmake/onnxruntime_providers_openvino.cmake:22

After raising the minimum supported OpenVINO version to 2026.0, the if(OpenVINO_VERSION VERSION_GREATER_EQUAL 2024.4) condition is always true and can be removed to avoid dead conditional logic.

  if(OpenVINO_VERSION VERSION_LESS 2026.0)
    message(FATAL_ERROR "OpenVINO 2026.0 and newer are supported. Please, use latest OpenVINO release")
  endif()

  if(OpenVINO_VERSION VERSION_GREATER_EQUAL 2024.4)
    add_definitions(-DUSE_OVEP_NPU_MEMORY=1)
  endif()

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

preetha-intel · 2026-06-30T09:25:10Z

+  // Allow NPU resources to be fully released between tests.
+  // Without this delay the NPU driver may fail to re-initialise
+  void TearDown() override {
+    std::this_thread::sleep_for(std::chrono::milliseconds(200));
+  }


+  static Ort::Session CreateSqueezeNetSession(
+      Ort::SessionOptions& session_options,
+      std::unordered_map<std::string, std::string>& ov_options) {
+    session_options.SetIntraOpNumThreads(1);


      - name: Set OpenVINORootDir
        shell: pwsh
        # Use $GITHUB_ENV to set the variable for subsequent steps
        run: |
-          $openVinoRootDir = Join-Path $env:RUNNER_TEMP "openvino-v2026.1.0"
+          $openVinoRootDir = Join-Path $env:RUNNER_TEMP "openvino-v2026.2.0"
          echo "OpenVINORootDir=$openVinoRootDir" >> $env:GITHUB_ENV


yuslepukhin · 2026-06-09T19:14:08Z

Issues Found

The .extension() == ".xml" check is case-sensitive. On Windows, .XML or .Xml would bypass this and fall through to GetOutputModelPath(). OpenVINO consistently produces lowercase .xml, so this is unlikely to trigger in practice, but a case-insensitive compare would be more robust.
std::filesystem::path(ep_cache_context) is constructed twice — once into cache_context_path and again as the second argument to ValidateExternalDataPath. Should reuse cache_context_path. Minor perf/readability issue, no correctness impact.
No test coverage for this fix. The workload-type test doesn't exercise the .xml vs non-.xml path resolution. A unit test with an .xml ep_cache_context that validates it resolves against the input model path (rather than output path) is missing.

MayureshV1 · 2026-06-09T22:44:39Z

@yuslepukhin.. This is a very targeted code path used by only Foundry customers and for models generated using Olive. This fix is to ensure the existing customer base is not impacted if using legacy EP.
Note, we are deprecating this path and moving our customer base to use plugin EP and hence don't have unit tests covering this scenario. However, there is internal validation for this code path which is how we caught the failing case.

If you still think we should add one please let us know and we can scope.

yuslepukhin · 2026-06-09T23:42:27Z

@yuslepukhin.. This is a very targeted code path used by only Foundry customers and for models generated using Olive. This fix is to ensure the existing customer base is not impacted if using legacy EP. Note, we are deprecating this path and moving our customer base to use plugin EP and hence don't have unit tests covering this scenario. However, there is internal validation for this code path which is how we caught the failing case.

If you still think we should add one please let us know and we can scope.

Regression test is a must for all PRs.

* Unit test for EPCTX OVIR model * [OVEP] Case-insensitive .xml check and reuse cache_context_path in EPCtxHandler * Address review comments

ankitm3k · 2026-06-23T12:44:18Z

@yuslepukhin Review comments have been addressed, please review & merge.

yuslepukhin · 2026-06-23T17:11:24Z

No negative test for malicious .xml path traversal: The OVIR tests validate the happy path. There's no test asserting that ../../../etc/evil.xml is rejected. The path validation utility covers this generically in its own tests (tensorutils_test.cc), but a targeted negative test for the OpenVINO EP's XML path would be ideal.
Workload type test TearDown() unconditional 200ms sleep (openvino_ep_workload_type_test.cc L36-38):
The Copilot reviewer correctly noted this adds unnecessary latency when tests are skipped. Both tests GTEST_SKIP() early if no NPU, but TearDown() still fires. This is wasteful and adds latencies to the build when re-used across the tests.

yuslepukhin · 2026-06-29T18:36:06Z

NEW BUG introduced in the latest update
The Windows CI workflow now has a version mismatch:

Download step uses OpenVINOVersion: 2026.2.1 → extracts to openvino-v2026.2.1
"Set OpenVINORootDir" step hardcodes "openvino-v2026.2.0" → points to a non-existent directory
This is exactly the fragility the Copilot reviewer warned about — the hardcoded path wasn't updated when the version was bumped from 2026.2.0 to 2026.2.1. The Set OpenVINORootDir step should either use the env var or be updated to "openvino-v2026.2.1".

Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com>

jatinwadhwa921 · 2026-06-30T10:27:48Z

@yuslepukhin Pls review

preetha-intel and others added 30 commits August 26, 2025 16:11

Update operator support status for OpenVINO 2025.2 (#792)

e019aa1

* Update operator support status for OpenVINO 2025.2 * Disable unsupported tests --------- Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com>

[OVEP] Fix cov issues (#796)

167f2ad

Merge branch 'master' into sync_msft_28082025

676a4d2

Merge pull request #797 from intel/sync_msft_28082025

b9a1885

Sync with Microsoft ONNX Runtime - 28/08/2025

remove onnxruntime/test/providers/openvino/openvino_plugin.cc (#798)

b7244f1

OVEP-CI updating version (#799)

be346bb

Revert "Sync with Microsoft ONNX Runtime - 01/09/2025 (#801)"

cc8b267

This reverts commit 2f1ad9d.

Merge pull request #803 from intel/revert-801-sync_msft_01092025

161e645

Revert "Sync with Microsoft ONNX Runtime - 01/09/2025"

Re-enable setting default precision for OV devices (#802)

c3276de

Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com>

bf16 tensor handling related fix (#805)

edc51ea

Handled tensors with bf16 data type and external in memory data Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com>

Merge branch 'master' into sync_msft_10_9_25

fba645c

Merge pull request #810 from intel/sync_msft_10_9_25

3215fef

Backmerging with Msft commits

[OVEP] Fix to increase provider value upto 2048 char (#807)

a5bd8ee

Make cache_dir and num_stream options session-local (#809)

02cf7e3

Co-authored-by: Preetha Veeramalai <preetha.veeramalai@intel.com>

Merge branch 'master' into sync_msft_15092025

f3d7b4f

Merge pull request #811 from intel/sync_msft_15092025

c098113

Sync with Microsoft ONNX Runtime - 15/09/2025

Merge branch 'master' into sync_msft_17092025

42b1fbb

Merge pull request #814 from intel/sync_msft_17092025

f7a5656

Sync with Microsoft ONNX Runtime - 17/09/2025

Merge branch 'master' into sync_msft_26092025

89db8a4

Merge pull request #818 from intel/sync_msft_26092025

9d56532

Sync with Microsoft ONNX Runtime - 26/09/2025

Merge branch 'master' into sync_msft_01102025

ce3a033

Merge pull request #820 from intel/sync_msft_01102025

2d82713

Sync with Microsoft ONNX Runtime - 01/10/2025

Merge branch 'master' into sync_msft_10102025

9a5e182

Merge pull request #826 from intel/sync_msft_10102025

2652479

Sync with Microsoft ONNX Runtime - 10/10/2025

When dynamic shapes are used the proto might be overriden and offsets…

b685871

… from initializers might become invalid (#828) Signed-off-by: bfilipek <bartlomiej.filipek@intel.com>

trigger stateful path for Phisilica model

65bbecc

Co-author: Beheshti, Nazanin

ankitm3k and others added 9 commits June 5, 2026 13:56

Merge pull request #1122 from intel/sync_msft_5_6_25

b76fc62

Backmerge pr Msft commits

Merge remote-tracking branch 'origin/master' into sync_msft_08062026

8605547

Update OV version to 2026.2.0

82ace8f

Merge pull request #1125 from intel/sync_msft_08062026

2aaf463

Sync with Microsoft ONNX Runtime - 08062026

Merge branch 'ovep-develop' into jatin_update_ov_2026_2

70741f6

Merge pull request #1126 from intel/jatin_update_ov_2026_2

421e511

Update OV version to 2026.2.0

Merge remote-tracking branch 'origin/master' into sync_msft_09062026

582bbc7

Merge pull request #1127 from intel/sync_msft_09062026

6f5465a

Sync with Microsoft ONNX Runtime - 09062026

fix & revert patch

2b6dcfa

[OVEP]Update Internal CI to OV 2026.2

762a9cf

yuslepukhin requested a review from Copilot June 9, 2026 19:05

Copilot started reviewing on behalf of yuslepukhin June 9, 2026 19:05 View session

Copilot AI reviewed Jun 9, 2026

View reviewed changes

ankitm3k and others added 3 commits June 10, 2026 15:50

Merge branch 'master' into ovep_dev_patch

cd68bda

Merge branch 'master' into ovep_dev_patch

6b21b7c

Unit test for EPCTX OVIR model (#1142)

c7d7df7

* Unit test for EPCTX OVIR model * [OVEP] Case-insensitive .xml check and reuse cache_context_path in EPCtxHandler * Address review comments

ankitm3k changed the title ~~[OVEP] OpenVINO Development Minor Updates~~ [OVEP] OpenVINO Development Updates Jun 23, 2026

Update Internal CI to OV 2026.2.1

5c028f2

jatinwadhwa921 force-pushed the ovep_dev_patch branch from 0b58ffd to 5c028f2 Compare June 30, 2026 01:53

jatinwadhwa921 and others added 3 commits June 30, 2026 03:04

Negative test case for EPCTX OVIR unit test case (#1165)

ab65671

Co-authored-by: MayureshV1 <47039074+MayureshV1@users.noreply.github.com>

Addressing review comments (#1171)

d2f9a36

Merge branch 'master' into ovep_dev_patch

f7196b6

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[OVEP] OpenVINO Development Updates#28954

[OVEP] OpenVINO Development Updates#28954
ankitm3k wants to merge 413 commits into
microsoft:mainfrom
intel:ovep_dev_patch

ankitm3k commented Jun 9, 2026 •

edited

Loading

Uh oh!

ankitm3k commented Jun 9, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

preetha-intel Jun 30, 2026

Uh oh!

yuslepukhin commented Jun 9, 2026

Uh oh!

MayureshV1 commented Jun 9, 2026 •

edited

Loading

Uh oh!

yuslepukhin commented Jun 9, 2026

Uh oh!

ankitm3k commented Jun 23, 2026

Uh oh!

yuslepukhin commented Jun 23, 2026 •

edited

Loading

Uh oh!

yuslepukhin commented Jun 29, 2026

Uh oh!

jatinwadhwa921 commented Jun 30, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

20 participants

Uh oh!

Conversation

ankitm3k commented Jun 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Uh oh!

ankitm3k commented Jun 9, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

preetha-intel Jun 30, 2026

Choose a reason for hiding this comment

Uh oh!

yuslepukhin commented Jun 9, 2026

Uh oh!

MayureshV1 commented Jun 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

yuslepukhin commented Jun 9, 2026

Uh oh!

ankitm3k commented Jun 23, 2026

Uh oh!

yuslepukhin commented Jun 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

yuslepukhin commented Jun 29, 2026

Uh oh!

jatinwadhwa921 commented Jun 30, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

20 participants

ankitm3k commented Jun 9, 2026 •

edited

Loading

MayureshV1 commented Jun 9, 2026 •

edited

Loading

yuslepukhin commented Jun 23, 2026 •

edited

Loading